AI-Enabled Wire-Arc Additive Manufacturing of TiB2-Modified Recycled AA7075 for Crack-Resistant Aerospace Structure

Abstract

The high-strength AA7075 aluminum alloy is attractive for aerospace and defense structures because of its favorable strength-to-weight ratio; however, its susceptibility to solidification cracking restricts its processing through wire-arc additive manufacturing. Simultaneously, the growing quantity of high-grade aluminum machining waste creates a need for direct and value-added recycling. This study developed a crack-resistant recycled AA7075 feedstock modified with in-situ TiB₂ grain refiners for wire-arc additive manufacturing of lightweight strategic components. Clean AA7075 machining chips were remelted, compositionally corrected, and converted into welding wires containing controlled TiB₂ additions. Wall structures were deposited by varying the wire-feed rate, travel speed, current, interpass temperature, and TiB- concentration. High-speed thermal imaging, arc-voltage signals, and acoustic emission data were recorded during deposition. A physics-informed machine-learning framework combining thermal-history constraints with XGBoost and Gaussian-process regression was used to predict the porosity, hot crack density, grain size, and mechanical anisotropy. The deposited structures were examined using X-ray computed tomography, optical microscopy, scanning electron microscopy, electron backscatter diffraction, and X-ray diffraction. Tensile, hardness, fracture toughness, and fatigue tests is performed to determine the relationship between process stability, TiB--induced heterogeneous nucleation, and structural performance. The life-cycle energy consumption and material recovery were also compared with those of conventionally produced AA7075. The principal novelty is the integration of recycled aerospace-grade aluminum, nanoparticle-assisted grain refinement, multimodal in situ sensing, and physically constrained defect prediction in a single additive-manufacturing framework. This study is expected to establish a closed-loop route for transforming aluminum waste into qualified, geometrically complex components while reducing hot cracking, process uncertainty, and dependence on primary aluminum.